This R01 renewal application extends our Community Based Participatory Research (CBPR) project that represents community efforts coupled with multi-institution, multi-disciplinary efforts across the perinatal period involving hospitals, WIC, prenatal and infant care providers to increase adoption of evidence-based approaches to improve breastfeeding (BF). Despite the proliferation of guidelines, recommendations and policies advising otherwise infants born to low income mothers (LIM) have higher rates of infant formula use and premature introduction of complementary food (CF), unsafe sleep conditions and second hand smoke (SHS) exposure compared to non-LIM. These rates represent significant current and future health care expenditures and are chronic and preventable and persistent. Our focus will expand beyond BF to include the concomitant issues of infant feeding practices (timing of CF), sleep safety and SHS exposure. These early infancy factors, while distinct, are inter-related, typically addressed to the same person (mother), in the same period of time (prenatal/early postpartum), are multifactorial, influenced by factors across levels of the social ecological (SE) model. Despite these commonalities, interventions promoting optimal infant feeding practices, safe sleep and SHS typically focus on a single issue at a single level (e.g. individual) and upon dissemination often fail to reach their original efficacy. Through this renewal we will continue to focus on two components of the SE model: Institutional level: address LIMs' lack of continuity, information and support across perinatal providers; Community level: CBPR and community messaging. While multi-level, multi-message coordinated interventions are considered essential to assure sustained impact, research testing their aggregate impact and across the SE model is lacking. Our long-term goal is to understand how key components of SE model, (intra/interpersonal, community and institutional) interact to learn how to maximally impact infant feeding practices and other infant health and safety practices. Our overarching hypothesis is that mothers will be more likely to follow guideline based infant feeding (and other health and safety practices) when all SE levels support these actions. Our CBPR approach and infrastructure uniquely positions us to pursue these aims: Increase community-wide adoption of evidence-based BF interventions across hospitals, WIC and prenatal and infant care providers to improve BF rates, delay introduction of CF and, secondarily increase adherence to SHS and safe sleep guidelines; test if a modified BFPC model to enhance continuity further improves these outcomes and; develop statistical models that account for the SE model's different levels (intra/interpersonal, community and institutional) investigating how these levels are associated with outcomes (BF, CF, SHS and safe sleep practices). Our results will: demonstrate how interventions across the SE model work synergistically (or do not) to increase adherence to feeding and safety guidelines and generate statistical models testing the SE model.